CN113450575B - Management method and device for roadside parking - Google Patents
Management method and device for roadside parking Download PDFInfo
- Publication number
- CN113450575B CN113450575B CN202110597859.5A CN202110597859A CN113450575B CN 113450575 B CN113450575 B CN 113450575B CN 202110597859 A CN202110597859 A CN 202110597859A CN 113450575 B CN113450575 B CN 113450575B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- frame
- image
- expression vector
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
Abstract
The embodiment of the invention provides a method and a device for managing roadside parking, wherein the method comprises the following steps: acquiring video image information acquired by a plurality of cameras, and determining a berthing area detection frame in each video frame; detecting whether vehicles exist in each video frame, if so, determining coordinate information of the berthing area detection frame, and extracting vehicle features of each image frame through a multilayer convolutional neural network to obtain a first feature expression vector of the vehicle; performing feature fusion on the first feature expression vector through a multilayer convolutional neural network to obtain a second feature expression vector, and calculating the motion vector direction and the motion distance of the vehicle in each frame of image; and analyzing the motion vector direction and the motion distance of the vehicle within a preset time according to the coordinate information of the parking area detection frame to obtain an analysis result and determine the parking event of the vehicle. By the method and the device, the entrance and exit events of the vehicles are accurately and efficiently identified.
Description
Technical Field
The invention relates to the technical field of intelligent parking management, in particular to a method and a device for managing roadside parking.
Background
The parking management based on high-level video has become an important subject in the construction and development of smart cities in recent years, and the parking management mode firstly carries out image video acquisition on vehicle and parking space information through a camera, and then carries out analysis and understanding on the vehicle information and vehicle behaviors through a computer vision technology, so that roadside parking is monitored and managed. However, the high-level video technology in the prior art is easily influenced by factors such as a complex scene, weather and object shielding, so that the accuracy cannot meet the application requirement of the scene; on the other hand, with the development of deep learning in recent years, deep learning based on a convolutional neural network has achieved significant achievement in the fields of image recognition, image detection, image segmentation and the like, but in the deep learning vehicle detection method in the prior art, detection is often performed by marking a vehicle with a detection frame, and if the background in an image acquired by a video is complicated, accurate image information cannot be recognized, so that the accuracy of the result of subsequent vehicle recognition, vehicle behavior analysis and the like is affected.
Therefore, a roadside parking management method which can efficiently and accurately detect the parked vehicles and has low management cost is urgently needed.
Disclosure of Invention
The embodiment of the invention provides a method and a device for managing roadside parking, which can accurately and efficiently identify an entrance and exit event of a vehicle.
In one aspect, an embodiment of the present invention provides a method for managing roadside parking, including:
acquiring video image information acquired by a plurality of cameras, and determining a berthing area detection frame in each video frame;
detecting whether vehicles exist in each video frame, if so, determining coordinate information of the berthing area detection frame, and extracting vehicle features of each image frame through a multilayer convolutional neural network to obtain a first feature expression vector of the vehicle;
performing feature fusion on the first feature expression vector through a multilayer convolutional neural network to obtain a second feature expression vector, and calculating the motion vector direction and the motion distance of the vehicle in each frame of image according to the second feature expression vector;
and analyzing the motion vector direction and the motion distance of the vehicle within a preset time length according to the coordinate information of the parking area detection frame to obtain an analysis result, and determining the parking event of the vehicle according to the analysis result.
Further, the acquiring video image information collected by a plurality of cameras and determining a parking area detection frame in each video frame includes:
determining coordinates of each vertex of the vehicle berth in a preset image acquisition area in each image;
and determining a berthage area detection frame in the video frame image acquired by each video camera according to the coordinates.
Further, the detecting whether a vehicle exists in each video frame, and if so, determining the coordinate information of the parking area detection frame includes:
detecting whether a vehicle exists in the image frames through a predetermined target detection algorithm;
and if so, determining the pixel coordinate information of each vertex of the parking area detection frame.
Further, the extracting the vehicle feature of each image frame through a multilayer convolutional neural network to obtain a first feature representation vector of the vehicle includes:
inputting the image frame into a multilayer convolution neural network according to the pixel coordinate information of each vertex of the berthage area detection frame;
and extracting and obtaining a first characteristic expression vector of the vehicle at a high layer and a first characteristic expression vector of a low layer in the plurality of convolutional layers through coding operation of the plurality of convolutional layers.
Further, the performing feature fusion on the first feature expression vector through a multilayer convolutional neural network to obtain a second feature expression vector includes:
stacking the extracted high-layer first feature expression vector and the low-layer first feature expression vector through a conversion layer of the multilayer convolutional neural network to obtain a stacked feature expression vector diagram;
and fusing the accumulated feature expression vector graphs to obtain a second feature expression vector.
Further, the calculating the motion vector direction and the motion distance of the vehicle in each frame of image according to the second feature expression vector includes:
and calculating the motion vector direction and the motion distance of the vehicle in each frame of image through dense optical flow calculation according to the second feature expression vector.
Further, the analyzing the motion vector direction and the motion distance of the vehicle within a predetermined time period according to the coordinate information of the parking area detection frame to obtain an analysis result, and determining the parking event of the vehicle according to the analysis result includes:
according to the coordinate information of the berth area detection frame, respectively determining the proportion of the number of video frames pointing to the inside of the berth area and the outside of the berth area in the motion vector direction of the vehicle within a preset time length and the total number of video frames within the preset time length in a two-dimensional coordinate system of any video frame image of each video image;
if the ratio of the number of video frames of which the motion vector direction of the vehicle points to the parking area in the preset time length to the total number of video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has an entrance event;
and if the ratio of the number of the video frames of which the motion vector directions of the vehicle point out of the berth area in the preset time length to the total number of the video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has a departure event.
On the other hand, an embodiment of the present invention provides a roadside parking management device, including:
the acquisition and determination module is used for acquiring video image information acquired by a plurality of cameras and determining a berth area detection frame in each video frame;
the extraction module is used for detecting whether vehicles exist in each video frame or not, if so, determining coordinate information of the berthing area detection frame, and extracting vehicle features of each image frame through a multilayer convolutional neural network to obtain a first feature expression vector of the vehicle;
the characteristic fusion and calculation module is used for carrying out characteristic fusion on the first characteristic expression vector through a multilayer convolutional neural network to obtain a second characteristic expression vector, and calculating the motion vector direction and the motion distance of the vehicle in each frame of image according to the second characteristic expression vector;
and the analysis and determination module is used for analyzing the motion vector direction and the motion distance of the vehicle within a preset time length according to the coordinate information of the parking area detection frame to obtain an analysis result, and determining the parking event of the vehicle according to the analysis result.
Further, the obtaining and determining module is specifically configured to
Determining coordinates of each vertex of the vehicle berth in a preset image acquisition area in each image;
and determining a berthage area detection frame in the video frame image acquired by each video camera according to the coordinates.
Further, the extraction module includes:
a detecting unit for detecting whether a vehicle exists in the image frames by a predetermined object detection algorithm;
and the determining unit is used for determining the pixel coordinate information of each vertex of the berth region detection frame if the pixel coordinate information exists.
Further, the extraction module includes:
the input unit is used for inputting the image frame to a multilayer convolution neural network according to the pixel coordinate information of each vertex of the berth area detection frame;
and an extraction unit for extracting and obtaining a high-level first feature expression vector and a low-level first feature expression vector of the vehicle in the plurality of convolutional layers through encoding operation of the plurality of convolutional layers.
Further, the feature fusion and calculation module includes:
a stacking unit, configured to stack the extracted high-level first feature representation vector and the low-level first feature representation vector through a conversion layer of the multilayer convolutional neural network to obtain a stacked feature representation vector diagram;
and the fusion unit is used for fusing the accumulated feature expression vector graphs to obtain a second feature expression vector.
Further, the feature fusion and calculation module includes:
and the calculating unit is used for calculating the motion vector direction and the motion distance of the vehicle in each frame of image through dense optical flow calculation according to the second feature expression vector.
Further, the analysis and determination module is specifically configured for
According to the coordinate information of the berth area detection frame, respectively determining the proportion of the number of video frames pointing to the inside of the berth area and the outside of the berth area in the motion vector direction of the vehicle within a preset time length and the total number of video frames within the preset time length in a two-dimensional coordinate system of any video frame image of each video image;
if the ratio of the number of video frames of which the motion vector direction of the vehicle points to the parking area in the preset time length to the total number of video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has an entrance event;
and if the ratio of the number of the video frames of which the motion vector directions of the vehicle point out of the berth area in the preset time length to the total number of the video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has a departure event.
The technical scheme has the following beneficial effects: according to the invention, the extracted vehicle characteristics are fused through the convolutional neural network, so that the data volume of the characteristics can be expanded, the accuracy of small target detection can be greatly improved, and the condition that the detection result has errors due to the fact that partial characteristics of the vehicle cannot be obtained due to the fact that the background of the acquired image is complex and the like is avoided; by accurately determining the motion vector direction and the motion vector distance of each frame of image, the accuracy rate of recognizing the vehicle exit/entrance events in the parking area can be greatly improved, meanwhile, the calculated amount can be greatly reduced, the recognition efficiency is greatly improved, and further, the parking management efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flowchart of a method for managing roadside parking in an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a roadside parking management device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The technical scheme of the embodiment of the invention has the following beneficial effects: according to the invention, the extracted vehicle characteristics are fused through the convolutional neural network, so that the data volume of the characteristics can be expanded, the accuracy of small target detection can be greatly improved, and the condition that the detection result has errors due to the fact that partial characteristics of the vehicle cannot be obtained due to the fact that the background of the acquired image is complex and the like is avoided; by accurately determining the motion vector direction and the motion vector distance of each frame of image, the accuracy rate of recognizing the vehicle exit/entrance events in the parking area can be greatly improved, meanwhile, the calculated amount can be greatly reduced, the recognition efficiency is greatly improved, and further, the parking management efficiency is improved.
The above technical solutions of the embodiments of the present invention are described in detail below with reference to application examples:
the application example of the invention aims to accurately and efficiently identify the entrance and exit events of the vehicle.
In a possible implementation manner, in the roadside parking management system, video images within a preset monitoring range are acquired through a plurality of video acquisition devices, then, video image information acquired by a plurality of cameras is acquired, and coordinates of each vertex of a vehicle berth in a preset image acquisition area in each image are determined; determining a parking area detection frame, such as a parking area detection frame A, in each video frame image acquired by each camera according to the coordinate information, then detecting whether a vehicle exists in the image frame through a preset target detection algorithm, if so, determining pixel coordinate information of each vertex of the parking area detection frame A, and extracting vehicle features of each image frame through a multilayer convolutional neural network to obtain a first feature expression vector of the vehicle C; performing feature fusion on the first feature expression vector through a multilayer convolutional neural network to obtain a second feature expression vector, and calculating the motion vector direction and the motion distance of the vehicle C in each frame of image according to the second feature expression vector; and analyzing the motion vector direction and the motion distance of the vehicle C within a preset time period, such as 5 minutes, according to the pixel coordinate information of each vertex of the parking area detection frame A to obtain an analysis result, and determining the parking event of the vehicle C according to the analysis result.
The berthing area detection frame comprises any one of a detection frame of each berthing area and detection frames of a plurality of berthing areas; the predetermined target detection algorithm comprises at least one of a single-stage, a two-stage and a non-anchor frame target detection algorithm; it should be noted that the area range of the detection frame of the parking area may be set according to the actual scene requirements.
In a possible implementation manner, performing vehicle feature extraction on each image frame through a multilayer convolutional neural network to obtain a first feature representation vector of the vehicle, including: inputting the image frame into a multilayer convolution neural network according to the pixel coordinate information of each vertex of the berthage area detection frame; and extracting and obtaining a first characteristic expression vector of the vehicle at a high layer and a first characteristic expression vector of a low layer in the plurality of convolutional layers through coding operation of the plurality of convolutional layers.
Performing feature fusion on the first feature expression vector through a multilayer convolutional neural network to obtain a second feature expression vector, including: stacking the extracted high-layer first feature expression vector and the low-layer first feature expression vector through a conversion layer of the multilayer convolutional neural network to obtain a stacked feature expression vector diagram; and fusing the accumulated feature expression vector graphs to obtain a second feature expression vector.
For example, in a roadside parking management system, video image information acquired by a plurality of cameras is acquired, and coordinates of each vertex of a vehicle berth in a predetermined image acquisition area in each image are determined; determining a berthage area detection frame, such as a berthage area detection frame A, in a video frame image acquired by each video camera according to the coordinate information, then detecting whether a vehicle exists in the image frame through a preset target detection algorithm, if so, determining pixel coordinate information of each vertex of the berthage area detection frame A, and then inputting the image frame into a multilayer convolutional neural network according to the pixel coordinate information of each vertex of the berthage area detection frame A; extracting and obtaining a high-level first feature expression vector and a low-level first feature expression vector of the vehicle in the plurality of convolutional layers through coding operation of the plurality of convolutional layers, and stacking the extracted high-level first feature expression vector and the low-level first feature expression vector through a conversion layer of the multilayer convolutional neural network to obtain a stacked feature expression vector diagram; and fusing the accumulated feature expression vector graphs to obtain a second feature expression vector.
According to the embodiment, the extracted high-level first feature expression vector and the low-level first feature expression vector are accumulated, so that the data amount of the features can be expanded, and the accumulated feature expression vector graphs are fused, so that the accuracy of small target detection can be greatly improved.
In a possible implementation manner, calculating a motion vector direction and a motion distance of the vehicle in each frame of image according to the second feature representation vector includes: and calculating the motion vector direction and the motion distance of the vehicle in each frame of image through dense optical flow calculation according to the second feature expression vector.
Analyzing the motion vector direction and the motion distance of the vehicle within a preset time according to the coordinate information of the parking area detection frame to obtain an analysis result, and determining a parking event of the vehicle according to the analysis result, wherein the method comprises the following steps: according to the coordinate information of the berth area detection frame, respectively determining the proportion of the number of video frames pointing to the inside of the berth area and the outside of the berth area in the motion vector direction of the vehicle within a preset time length and the total number of video frames within the preset time length in a two-dimensional coordinate system of any video frame image of each video image; if the ratio of the number of video frames of which the motion vector direction of the vehicle points to the parking area in the preset time length to the total number of video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has an entrance event; and if the ratio of the number of the video frames of which the motion vector directions of the vehicle point out of the berth area in the preset time length to the total number of the video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has a departure event.
For example, in the roadside parking management system, after the second feature expression vector is obtained as described above, the motion vector direction and the motion distance of the vehicle C in each frame of image are calculated through dense optical flow calculation according to the obtained second feature expression vector, and the ratio of the number of video frames pointing into the parking area and outside the parking area to the total number of video frames within 5 minutes in a predetermined time period, such as within 5 minutes, in the two-dimensional coordinate system of any video frame image of each video image is respectively determined according to the pixel coordinate information of each vertex of the parking area detection frame a; if the ratio of the number of the video frames of which the motion vector direction of the vehicle A points to the parking area within 5 minutes to the total video frame number within 5 minutes is greater than a preset threshold value, determining that the vehicle A has an entrance event; and if the ratio of the number of the video frames of which the motion vector direction of the vehicle A points out of the berth area within 5 minutes to the total number of the video frames within 5 minutes is greater than a preset threshold value, determining that the departure event of the vehicle A occurs.
The embodiment of the present invention provides a roadside parking management device, which can implement the method embodiments provided above, and for specific function implementation, reference is made to the description in the method embodiments, and details are not repeated here.
It should be understood that the specific order or hierarchy of steps in the processes disclosed is an example of exemplary approaches. Based upon design preferences, it is understood that the specific order or hierarchy of steps in the processes may be rearranged without departing from the scope of the present disclosure. The accompanying method claims present elements of the various steps in a sample order, and are not intended to be limited to the specific order or hierarchy presented.
In the foregoing detailed description, various features are grouped together in a single embodiment for the purpose of streamlining the disclosure. This method of disclosure is not to be interpreted as reflecting an intention that the claimed embodiments of the subject matter require more features than are expressly recited in each claim. Rather, as the following claims reflect, invention lies in less than all features of a single disclosed embodiment. Thus, the following claims are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. To those skilled in the art; various modifications to these embodiments will be readily apparent, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the disclosure. Thus, the present disclosure is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
What has been described above includes examples of one or more embodiments. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the aforementioned embodiments, but one of ordinary skill in the art may recognize that many further combinations and permutations of various embodiments are possible. Accordingly, the embodiments described herein are intended to embrace all such alterations, modifications and variations that fall within the scope of the appended claims. Furthermore, to the extent that the term "includes" is used in either the detailed description or the claims, such term is intended to be inclusive in a manner similar to the term "comprising" as "comprising" is interpreted when employed as a transitional word in a claim. Furthermore, any use of the term "or" in the specification of the claims is intended to mean a "non-exclusive or".
Those of skill in the art will further appreciate that the various illustrative logical blocks, units, and steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate the interchangeability of hardware and software, various illustrative components, elements, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design requirements of the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present embodiments.
The various illustrative logical blocks, or elements, described in connection with the embodiments disclosed herein may be implemented or performed with a general purpose processor, a digital signal processor, an Application Specific Integrated Circuit (ASIC), a field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general-purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a digital signal processor and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a digital signal processor core, or any other similar configuration.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. For example, a storage medium may be coupled to the processor such the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an ASIC, which may be located in a user terminal. In the alternative, the processor and the storage medium may reside in different components in a user terminal.
In one or more exemplary designs, the functions described above in connection with the embodiments of the invention may be implemented in hardware, software, firmware, or any combination of the three. If implemented in software, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media that facilitate transfer of a computer program from one place to another. Storage media may be any available media that can be accessed by a general purpose or special purpose computer. For example, such computer-readable media can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to carry or store program code in the form of instructions or data structures and which can be read by a general-purpose or special-purpose computer, or a general-purpose or special-purpose processor. Additionally, any connection is properly termed a computer-readable medium, and, thus, is included if the software is transmitted from a website, server, or other remote source via a coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL), or wirelessly, e.g., infrared, radio, and microwave. Such discs (disk) and disks (disc) include compact disks, laser disks, optical disks, DVDs, floppy disks and blu-ray disks where disks usually reproduce data magnetically, while disks usually reproduce data optically with lasers. Combinations of the above may also be included in the computer-readable medium.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (12)
1. A method of managing roadside parking, comprising:
acquiring video image information acquired by a plurality of cameras, and determining a berthing area detection frame in each video frame;
detecting whether vehicles exist in each video frame, if so, determining coordinate information of the berthing area detection frame, and extracting vehicle features of each image frame through a multilayer convolutional neural network to obtain a first feature expression vector of the vehicle;
performing feature fusion on the first feature expression vector through a multilayer convolutional neural network to obtain a second feature expression vector, and calculating the motion vector direction and the motion distance of the vehicle in each frame of image according to the second feature expression vector;
analyzing the motion vector direction and the motion distance of the vehicle within a preset time length according to the coordinate information of the parking area detection frame to obtain an analysis result, and determining a parking event of the vehicle according to the analysis result;
the step of analyzing the motion vector direction and the motion distance of the vehicle within a preset time period according to the coordinate information of the parking area detection frame to obtain an analysis result, and determining the parking event of the vehicle according to the analysis result comprises the following steps:
according to the coordinate information of the berth area detection frame, respectively determining the proportion of the number of video frames pointing to the inside of the berth area and the outside of the berth area in the motion vector direction of the vehicle within a preset time length and the total number of video frames within the preset time length in a two-dimensional coordinate system of any video frame image of each video image;
if the ratio of the number of video frames of which the motion vector direction of the vehicle points to the parking area in the preset time length to the total number of video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has an entrance event;
and if the ratio of the number of the video frames of which the motion vector directions of the vehicle point out of the berth area in the preset time length to the total number of the video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has a departure event.
2. The method of claim 1, wherein the obtaining video image information captured by a plurality of cameras and determining the berthage area detection frame in each video frame comprises:
determining coordinates of each vertex of the vehicle berth in a preset image acquisition area in each image;
and determining a berthage area detection frame in the video frame image acquired by each video camera according to the coordinates.
3. The method of claim 2, wherein the detecting whether a vehicle exists in each video frame and, if so, determining coordinate information of the parking area detection frame comprises:
detecting whether a vehicle exists in the image frames through a predetermined target detection algorithm;
and if so, determining the pixel coordinate information of each vertex of the parking area detection frame.
4. The method of claim 3, wherein the extracting vehicle features from each image frame through a multilayer convolutional neural network to obtain a first feature representation vector of the vehicle comprises:
inputting the image frame into a multilayer convolution neural network according to the pixel coordinate information of each vertex of the berthage area detection frame;
and extracting and obtaining a first characteristic expression vector of the vehicle at a high layer and a first characteristic expression vector of a low layer in the plurality of convolutional layers through coding operation of the plurality of convolutional layers.
5. The method of claim 4, wherein the feature fusing the first feature representation vector through a multilayer convolutional neural network to obtain a second feature representation vector, comprises:
stacking the extracted high-layer first feature expression vector and the low-layer first feature expression vector through a conversion layer of the multilayer convolutional neural network to obtain a stacked feature expression vector diagram;
and fusing the accumulated feature expression vector graphs to obtain a second feature expression vector.
6. The method according to claim 5, wherein the calculating of the motion vector direction and the motion distance of the vehicle in each frame of image according to the second feature representation vector comprises:
and calculating the motion vector direction and the motion distance of the vehicle in each frame of image through dense optical flow calculation according to the second feature expression vector.
7. A roadside parking management device, comprising:
the acquisition and determination module is used for acquiring video image information acquired by a plurality of cameras and determining a berth area detection frame in each video frame;
the extraction module is used for detecting whether vehicles exist in each video frame or not, if so, determining coordinate information of the berthing area detection frame, and extracting vehicle features of each image frame through a multilayer convolutional neural network to obtain a first feature expression vector of the vehicle;
the characteristic fusion and calculation module is used for carrying out characteristic fusion on the first characteristic expression vector through a multilayer convolutional neural network to obtain a second characteristic expression vector, and calculating the motion vector direction and the motion distance of the vehicle in each frame of image according to the second characteristic expression vector;
the analysis and determination module is used for analyzing the motion vector direction and the motion distance of the vehicle within a preset time length according to the coordinate information of the parking area detection frame to obtain an analysis result, and determining a parking event of the vehicle according to the analysis result;
the analysis and determination module is also used for
According to the coordinate information of the berth area detection frame, respectively determining the proportion of the number of video frames pointing to the inside of the berth area and the outside of the berth area in the motion vector direction of the vehicle within a preset time length and the total number of video frames within the preset time length in a two-dimensional coordinate system of any video frame image of each video image;
if the ratio of the number of video frames of which the motion vector direction of the vehicle points to the parking area in the preset time length to the total number of video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has an entrance event;
and if the ratio of the number of the video frames of which the motion vector directions of the vehicle point out of the berth area in the preset time length to the total number of the video frames in the preset time length is greater than a preset threshold value, determining that the vehicle has a departure event.
8. The device according to claim 7, wherein the acquisition and determination module is specifically configured to
Determining coordinates of each vertex of the vehicle berth in a preset image acquisition area in each image;
and determining a berthage area detection frame in the video frame image acquired by each video camera according to the coordinates.
9. The apparatus of claim 8, wherein the extraction module comprises:
a detecting unit for detecting whether a vehicle exists in the image frames by a predetermined object detection algorithm;
and the determining unit is used for determining the pixel coordinate information of each vertex of the berth region detection frame if the pixel coordinate information exists.
10. The apparatus of claim 9, wherein the extraction module comprises:
the input unit is used for inputting the image frame to a multilayer convolution neural network according to the pixel coordinate information of each vertex of the berth area detection frame;
and an extraction unit for extracting and obtaining a high-level first feature expression vector and a low-level first feature expression vector of the vehicle in the plurality of convolutional layers through encoding operation of the plurality of convolutional layers.
11. The apparatus of claim 10, wherein the feature fusion and computation module comprises:
a stacking unit, configured to stack the extracted high-level first feature representation vector and the low-level first feature representation vector through a conversion layer of the multilayer convolutional neural network to obtain a stacked feature representation vector diagram;
and the fusion unit is used for fusing the accumulated feature expression vector graphs to obtain a second feature expression vector.
12. The apparatus of claim 11, wherein the feature fusion and computation module comprises:
and the calculating unit is used for calculating the motion vector direction and the motion distance of the vehicle in each frame of image through dense optical flow calculation according to the second feature expression vector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110597859.5A CN113450575B (en) | 2021-05-31 | 2021-05-31 | Management method and device for roadside parking |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110597859.5A CN113450575B (en) | 2021-05-31 | 2021-05-31 | Management method and device for roadside parking |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113450575A CN113450575A (en) | 2021-09-28 |
| CN113450575B true CN113450575B (en) | 2022-04-19 |
Family
ID=77810428
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110597859.5A Active CN113450575B (en) | 2021-05-31 | 2021-05-31 | Management method and device for roadside parking |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113450575B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114530056B (en) * | 2022-02-15 | 2023-05-02 | 超级视线科技有限公司 | Parking management method and system based on positioning information and image information |
| CN115035741B (en) * | 2022-04-29 | 2024-03-22 | 阿里云计算有限公司 | Method, device, storage medium and system for discriminating parking position and parking |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108305466B (en) * | 2018-03-13 | 2020-05-08 | 北京智芯原动科技有限公司 | Roadside parking detection method and device based on vehicle characteristic analysis |
| CN110533955B (en) * | 2019-09-16 | 2022-12-09 | 平安科技（深圳）有限公司 | Method for determining parking space, terminal equipment and computer readable storage medium |
| CN111476169B (en) * | 2020-04-08 | 2023-11-07 | 智慧互通科技股份有限公司 | Complex scene road side parking behavior identification method based on video frame |
| CN111611886B (en) * | 2020-04-24 | 2023-11-10 | 智慧互通科技股份有限公司 | Parking information verification method and device based on vehicle behavior analysis |
| CN111739335B (en) * | 2020-04-26 | 2021-06-25 | 智慧互通科技股份有限公司 | Parking detection method and device based on visual difference |
| CN111986508A (en) * | 2020-08-24 | 2020-11-24 | 广州信息投资有限公司 | Roadside parking management method and system based on multi-target tracking and visual positioning |
| CN112258668A (en) * | 2020-10-29 | 2021-01-22 | 成都恒创新星科技有限公司 | Method for detecting roadside vehicle parking behavior based on high-position camera |
| CN112766222B (en) * | 2020-12-30 | 2024-04-09 | 爱泊车科技有限公司 | Method and device for assisting in identifying vehicle behavior based on berth line |
-
2021
- 2021-05-31 CN CN202110597859.5A patent/CN113450575B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113450575A (en) | 2021-09-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110491168B (en) | Method and device for detecting vehicle parking state based on wheel landing position | |
| CN110163107B (en) | Method and device for recognizing roadside parking behavior based on video frames | |
| CN113450575B (en) | Management method and device for roadside parking | |
| CN111739175B (en) | Vehicle parking monitoring method and device | |
| CN111739335B (en) | Parking detection method and device based on visual difference | |
| WO2021031954A1 (en) | Object quantity determination method and apparatus, and storage medium and electronic device | |
| CN110688902B (en) | Method and device for detecting vehicle area in parking space | |
| CN113055823B (en) | Method and device for managing shared bicycle based on road side parking | |
| CN113205692B (en) | Automatic identification method for road side parking position abnormal change | |
| CN111178235A (en) | Target quantity determination method, device, equipment and storage medium | |
| CN114049356B (en) | Method, device and system for detecting structure apparent crack | |
| CN111739338A (en) | Parking management method and system based on multiple types of sensors | |
| CN113205691A (en) | Method and device for identifying vehicle position | |
| CN112836699A (en) | Long-time multi-target tracking-based berth entrance and exit event analysis method | |
| CN114997307A (en) | Trajectory prediction method, apparatus, device and storage medium | |
| CN112766222B (en) | Method and device for assisting in identifying vehicle behavior based on berth line | |
| CN113869258A (en) | Traffic incident detection method and device, electronic equipment and readable storage medium | |
| CN113066306B (en) | Management method and device for roadside parking | |
| CN111951601B (en) | Method and device for identifying parking positions of distribution vehicles | |
| CN113052141A (en) | Method and device for detecting parking position of vehicle | |
| KR102260556B1 (en) | Deep learning-based parking slot detection method and apparatus integrating global and local information | |
| CN116612103B (en) | Intelligent detection method and system for building structure cracks based on machine vision | |
| CN114463976A (en) | Vehicle behavior state determination method and system based on 3D vehicle track | |
| CN113408514A (en) | Method and device for detecting roadside parking lot berth based on deep learning | |
| Noor et al. | Automatic parking slot occupancy detection using Laplacian operator and morphological kernel dilation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |